Progress on weed biocontrol projects
in Paraná State, Brazil: targeting plants
that are invasive in Brazil and
elsewhere in the world
J.H. Pedrosa-Macedo1
Summary
Biological control projects against weeds in Paraná State, Brazil are currently focused on six species
or species groups. Two projects, including native Senecio species (Asteraceae) and the exotic Tecoma
stans (L.) Kunth (Bignoniaceae), involve studies on the ecology and natural enemies of plants that are
problematic in Brazil. Native Senecio species are particularly problematic in Brazilian pastures, where
losses of livestock, notably cattle, due to toxicity have been estimated to cost US$7.5 million annually
in the State of Rio Grande do Sul alone. The Central American T. stans, which has become an
increasing problem in pastures in Paraná State, is currently the subject of botanical studies and studies
on associated insects and fungal pathogens. The remaining four projects represent cooperative research
programs with international organisations and involve studies on the ecology and natural enemies of
native Brazilian plants that are invasive elsewhere in the world. These species include Schinus terebinthifolius Raddi (Anacardiaceae) and Tibouchina herbacea (Melastomataceae), both of which are problematic in Hawaii, Psidium cattleianum Sabine (Myrtaceae), which is invasive in Florida (USA), and
Solanum mauritianum Scopoli (Solanaceae), which invades the high rainfall regions of South Africa.
This paper updates the progress achieved with these six projects.
Keywords: Psidium cattleianum, Schinus terebinthifolius, Senecio spp., Solanum
mauritianum, Tecoma stans, Tibouchina herbacea.
Introduction
The Federal University, Paraná, Brazil (UFPR), has
encouraged studies on agricultural weeds as well as the
training of weed specialists in the departments of Forest
Sciences, Veterinary Studies and Entomology since
1990. These programs led to the establishment of international programs aimed principally at the development of biocontrol agents against Brazilian species that
have become pests elsewhere.
Several species of Senecio kill cattle in Rio Grande
do Sul State (Riet-Correa et al. 1991). S. brasiliensis is
the most problematic species, as it is a significant weed
in pastures in the southern states of Brazil. It is a large,
1
Centro de Ciências Florestais – DECIF-SCA – Universidade Federal do
Paraná, Av. Lothário Meissner, 3400, 80.210-170 Curitiba, Paraná,
Brazil. Laboratório Neotropical de Controle Biológico de Plantas
<[email protected]>.
perennial plant that can form significant stands within
2–3 years. The economic losses due to intoxication
induced by Senecio species are estimated at US$7.5
million (Mendez 1997, Riet-Correa & Medeiros 2000,
Karam et al. 2002). Our intention is to augment populations of the most damaging insects in the hope that
they will reduce stands significantly.
The “amarelinho” (yellow bells), Tecoma stans, is
another priority species. This plant, which has spread
throughout several parts of the world, presents conflicts
of interest in that it is considered either as a weed, an
ornamental or a medicinal plant (Kranz & Passini
1997). It is considered an invasive alien plant in the
Brazilian savanna (Mendonça et al. 1998).
International cooperative studies are being
conducted against strawberry guava (Psidium
cattleianum – Myrtaceae), Brazilian peppertree
(Schinus terebinthifolius – Anacardiaceae), glory bush
310
Weed biocontrol in Paraná
(Tibouchina herbacea – Melastomataceae) and
bugweed (Solanum mauritianum – Solanaceae). There
is a wide range of insects that cause damage on leaves,
growth buds, flower buds, flowers and fruits of these
plants. Over 18 species are undergoing intensive field
and laboratory research aimed at elucidating their
biology, host specificity and impact. To facilitate field
and laboratory studies, an arboretum has been developed containing a variety of species in the following
families: Asteraceae, Anacardiaceae, Bignoniaceae,
Convolvulaceae,
Lauraceae,
Melastomataceae,
Poaceae, Solanaceae, Myrtaceae and Rutaceae.
Research is carried out in collaboration with various
UFPR units. FUPEF (Fundação de Pesquisas Florestais
do Paraná) is a university foundation that administers
most foreign cooperative agreements (currently
University of Hawaii, University of Florida, and Plant
Protection Institute of South Africa). Besides these
agreements, there are cooperative arrangements with
local universities and government agencies, principally
UNICENTRO (Universidade Estadual do CentroOeste, Irati, PR), FURB (Fundação Universidade de
Blumenau, SC), UEL (Universidade Estadual de Londrina, PR), and UFPel (Universidade Federal de Pelotas,
RS).
first plateau ranges from 650 to 1100 m. The climate is
hot and humid with temperatures between 15 and 19°C
(min. –10°C, max. 35°C), 0–40 frosts per year and
900–1200 mm annual rainfall. Droughts are extremely
rare (Carpanezzi et. al. 1986). Within the region there
are several protected forest areas, including Araucaria
angustifolia (Araucariaceae) cloud forest, as well as
extensive secondary forest and submontane fields
available for studies.
The study region
Tecoma stans (Bignoniaceae) – amarelinho
– yellow bells
Universidade Federal do Paraná has an ideal location in
Curitiba due to its proximity to several major habitats,
viz. Atlantic Forest (tropical, from sea level to 650 m),
restinga (sea level to 20 m) and Araucaria Forest
(subtropical – 650 to 1100 m). All are within 90–100
km of the university. This closeness enables studies of
insects and diseases associated with the target plants to
be undertaken throughout the year under natural conditions, as well as in the greenhouses and arboreta. The
conditions at Curitiba were advantageous enough that
CSIRO (Australia) maintained a biological control
laboratory there during the 1970s and 1980s. Coastal
restinga vegetation consists of a wide variety of plants
growing in sandy substrate with a high water table. The
forest today is all secondary and of short stature, not
more than 5 m. There are a few protected areas, but the
region has marginal agricultural potential so disturbance is minimal. The climate is hot and humid with
year-round average temperature between 18–22°C
(min. 0.9°C, max. 38°C) and annual rainfall between
1430 and 2450 mm (Maack 1968). Droughts are
possible in June and July (Carpanezzi et al. 1986). The
Atlantic Forest ranges from 50–700 m. It is mostly
secondary forest with some pockets of primary vegetation. The climate is humid, tropical, with temperatures
from 15 to 19°C (min. 5°C, max. 38°C) and annual
rainfall between 1250 and 2500 mm. In Paraná, much
of this forest is protected. However, some limited
subsistence agriculture is practised on a small scale in
some areas (Carpanezzi et al. 1986). The subtropical
Biological control research projects
The current research program at Curitiba includes
efforts to control Tecoma stans in Paraná State and
Senecio brasiliensis (Asteraceae) in southern Brazil
(Paraná, Santa Catarina and Rio Grande do Sul). The
cooperative program with the University of Florida
(Brazilian peppertree control project) and with the
University of Hawaii (strawberry guava and
Tibouchina biological control projects) are continuing
at an advanced phase of agent selection, host range and
impact tests. The cooperative program with the Plant
Protection Research Institute, South Africa is also
active, testing the host range of one only agent on six
Solanaceae species.
The plant is found throughout much of Brazil, from
Amazonia (Manaus) south to Rio Grande do Sul
(Butiá). In the north and northeastern regions of the
Paraná state, it occupies over 50,000 ha of pastureland.
The infestation is centered around Londrina, where
over 10,000 ha of pasture has been lost to weed infestations. Yet, in Curitiba and vicinities, it is used as in
urban forestry from where it has dispersed to abandoned grounds.
We have initiated phenological studies. A number of
insects have been collected, including Lepidoptera (2
spp.), Coleoptera (9 spp.), Homoptera (2 spp.),
Hymenoptera (6 spp.), Hemiptera (5 spp.) and Thysanoptera (1 sp.). Two unidentified Lepidoptera species
attack the plant: a leaf-roller (Crambidae) and a fruit
borer (Olethreutidae). The other associated insects are
being collected for future taxonomic and biology
studies. At Bogotá savanna (Colombia), there are 41
associated insect species: Coleoptera (1 sp.), Diptera (3
spp.), Hemiptera (6 spp.), Homoptera (9 spp.),
Hymenoptera (14 spp.), Lepidoptera (3 spp.), Neuroptera (2 spp.) and Thysanoptera (3 spp.). In addition,
there are also three Acari species (Lee et al. 2000). In
Nicaragua (Masaya National Park), the species
observed to date include: Lepidoptera (1 sp.), Coleoptera (3 spp.), Homoptera (2 spp.) and Hymenoptera (3
spp) (Pedrosa-Macedo, personal notes, 2002). At
Blumenau, SC (FURB) studies on associated insects
and fungal diseases are being conducted (Prospodium
311
Proceedings of the XI International Symposium on Biological Control of Weeds
appendiculatum), while at Londrina (UEL), the studies
include botanical and entomological aspects.
somewhat incongruous that one form of a currently
accepted species is not attacked by an insect, whereas
another species is.
Senecio (Asteraceae) in Brazil
There are about 150 species of Senecio in Brazil, but
the biennial S. heterotrichius, S. selloi, S. brasiliensis
and the annual S. oxyphyllus are the most common in
pastures (Karam et al. 2002). The most abundant potential control agents are: Pericopis sacrifica (Lepidoptera), Phaedon confinis, Agathomerus subfasciatus and
Systena s-littera tenus (Coleoptera) (Pedrosa-Macedo
et al. 2000, Karam et al. 2002). All these species are
native to the regions of infestation. Thus, their use in an
augmentation biological control program is limited
because their natural enemies are also present. Potential
control agents from other areas, including Longitarsus
jacobaeae Waterhouse and L. flavicornis Stephens
(Coleoptera: Chrysomelidae), could be used as a “short
cut” project, as both are already used as biological
control agents for Senecio jacobaea L. (Asteraceae)
(tansy ragwort). Further evaluation depends on international cooperation.
Psidium cattleianum (Myrtaceae) –
strawberry guava
Strawberry guava is established in at least 31 countries (mostly islands) in the subtropical region (Wikler
1999). In the Hawaiian archipelago it is a significant
weed of native forest, where it forms monotypic stands.
Ecosystem disturbance, particularly by non-indigenous
feral pigs, is the principal mechanism of establishment
and intensification (Diong 1982). Manual control,
though expensive, is feasible, but the ecological and
often archaeological damage is unacceptable. Chemical
control is increasingly difficult as more and more suitable herbicides become banned due to long-term undesirable ecological effects. Biological control is the last
resort. Potential biological control agents have been
discovered in the Atlantic Forest and associated areas in
Brazil. They include, in order of suitability: Tectococcus ovatus (Homoptera: Eriococcidae); a leaf galler,
Dasineura gigantea (Hymenoptera: Cecidomyiidae); a
bud galler, Eurytoma psidii; and other gall-formers,
either Eurytoma cattleianii or Eurytoma desantisi
(Hymenoptera, Eurytomidae) (Angelo 1997, Vitorino
2001). Neotrioza tavaresi (Hemiptera: Psyllidae),
another leaf galler, appears to have insufficient impact
on the plants to be useful (Butignol & Pedrosa-Macedo
2001). None of the above species attack the congeneric
P. guajava, an important agricultural fruit crop.
The taxonomy of the P. cattleianum group needs
further research. Field tests demonstrated that the
yellow-fruited form of Psidium cattleianum and P.
spathulatum are heavily attacked by T. ovatus, whereas
the red-fruited form of P. cattleianum and P. longipetiolatum appear to be resistant. Insects are frequently
extremely capable discriminators between species. It is
Solanum mauritianum (Solanaceae) –
bugweed (fumo-bravo)
This plant was taken to South Africa by Portuguese
navigators in the 16th century (Roe 1972, Olckers
1999). It is a significant weed in reafforestation, agriculture and conservation areas, urban space, river
margins and road margins. It is listed as a Category I
weed in the South African biological control program
(Henderson 2001). The weevil Anthonomus morticinus
Clark (Coleoptera: Curculionidae) from south-eastern
Brazil is being studied on nine congeners: S. capsicoides, S. diflorum, S. fastigiatum, S. gilo, S. granulosoleprosum, Solanum melongena, S. palinacantum, S.
tuberosum, and S. viarum. Field studies are conducted
at various sites around Curitiba and laboratory and
controlled environment studies at the Arboretum
Juvevê. None of the above Solanum species is attacked
by A. morticinus, except S. mauritianum and S. granuloso-leprosum. In “no-choice” tests in the laboratory,
however, this weevil can feed on some of these species.
Schinus terebinthifolius (Anacardiaceae) –
Brazilian peppertree (aroeira)
Brazilian peppertree is an aggressive plant in
Florida, USA, especially in the Everglades National
Park, and also in the Hawaiian archipelago. It was
brought to the USA as an ornamental plant in 1840
(Bennet & Habeck 1991) and again in 1891 (Workman
1978). It is native to southern Brazil, Paraguay and
northern Argentina. It has been spread subsequently to
several parts of the world, including American Samoa,
Australia, Fiji, New Caledonia, Mauritius, Micronesia,
Puerto Rico, and Tahiti. Four potential biological
control agents are being studied: Heteroperreyia
hubrichi (Hymenoptera: Pergidae), Calophya terebinthifolii (Homoptera: Psyllidae), Epsimus utilis (Lepidoptera: Tortricidae) and Pseudophilothrips ichini
(Thysanoptera: Phlaeothripidae). H. hubrichi was
approved after host range tests in Florida, but not in
Hawaii where it also attacks the endemic Rhus hawaiiensis. Its toxicity to wild animals and cattle is being
evaluated because there are some suggestive reports
from Australia. A preliminary test with a calf, where
100 final-instar larvae were mixed with the food,
resulted in no signs of poisoning after 24 and 72 hours.
The droughts associated with the “El Niño” phenomenon have resulted in a significant reduction in H.
hubrichi populations, curtailing the program temporarily. We have not developed a reliable mass-rearing
technology to date.
The leaflet galler, C. terebinthifolii, has a disjunct
distribution on the First Plateau of Paraná and the
littoral area. There are different impacts in the two
312
Weed biocontrol in Paraná
regions. Both of these results have caused some
concern as to the suitability of this potential agent.
Further studies on its biology, as well as captive rearing
studies, are in progress.
Epsimus utilis provides a reasonable degree of
specificity to the Brazilian peppertree. Preliminary tests
with E. utilis larvae on young Brazilian peppertree
plants resulted in a mean biomass loss of 42% per
leaflet, but the general impact of the larvae on the plant
was not established; we have not developed a sufficiently robust means of evaluation.
Pseudophilothrips ichini attacks Brazilian peppertree throughout its range, except in the littoral areas of
Paraná, where it is rare. “No-choice” tests registered
attacks on Mangifera indica, Anacardium ocidentale
and Rhus sandwicensis (Anacardiaceae), but in
“multiple choice” tests these species were not attacked.
Two types of impact have been established: 11% reduction of plant growth and 45% increase in branching,
causing excessive branching of the plant. P. ichini tests
on the ornamental Schinus molle are being prepared.
ment District, University of Florida, Institute of Food
and Agricultural Sciences; Plant Protection Research
Institute of Pietermarizburg, South Africa; Ministério
do Meio Ambiente (MMA); Conselho Nacional de
Desenvolvimento Científico e Tecnológico (CNPq);
and FUPEF-Fundação de Pesquisas Florestais do
Paraná-Curitiba, Brazil, for financial and administration support. Thanks are also due to Clifford W. Smith,
Dale H. Habeck, Stephen Hight, James P. Cuda, Julio
Medal, Terry Olckers, Matheus Tracy Johnson,
Marcelo Diniz Vitorino, Ayeres de Oliveira Menezes
Jr., Charles Wikler, Cesar A. Butignol, Germano
Henrique Rosado Neto, João Ricardo Dittrich, Lúcia
Massutti de Almeida, Anamaria Dal Molin, Cecília
Gonçalves Simões, Dalila Aparecida Harmuch, Deise
Mari Barboza, Kelly Hacke Ribeiro, Lorena Stolle,
Luizimir Eduardo Furmann, Marcelo Galeazzi
Caxambu, Márcia Cristina Mendes Marques, Marcelo
Mattos de Paula and Zildo Luiz Ramos for their considerable assistance.
References
Tibouchina herbacea
T. herbacea was introduced into Hawaii as an ornamental plant. It is originally from Brazil, and the main
population occurs at Serra Gaucha, Rio Grande do Sul
state, although it is known from other Brazilian states
such as Santa Catarina and in the cerrado (Brazilian
savanna) region (Mendonça et al. 1998). It belongs to a
species complex whose systematics is unclear. Two
closely related species, Tibouchina cerastifolia and
T. gracilis, are common. Two other related species,
Acisanthera variabilis and Rhyncanthera sp., are also
frequent. All four species are attacked by at least three
different species of as yet unidentified weevils. Among
35 insect species associated with ruderal herbaceous
Melastomataceae, six are potential agents: Schrenkensteinia sp. (Lepidoptera: Schrenkensteinidae), whose
presence was verified only in Rio Grande do Sul state,
though Barreto (pers. comm.) recorded it in Minas
Gerais; Syphrea uberabensis Bechyné (Coleoptera:
Chrysomelidae); Margaridisa sp. (Coleoptera: Chrysomelidae); Lius sp. (Coleoptera: Buprestidae); an
unidentified Geometridae caterpillar (Lepidoptera);
and Anthonomus partiarius Boheman (Coleoptera:
Curculionidae). The last insect was found on
T. herbacea, T. cerastifolia and Acisanthera variabilis,
feeding on flower buds, flowers, pollen and seeds.
Studies on its biology, ecological behaviour and host
range are being conducted. A. partiarius was not found
on T. gracilis and Rhyncanthera sp. (Pedrosa-Macedo
et al. 2000)
Acknowledgements
Thanks are due to: the Cooperative National Park
Resources Studies Unit, University of Hawaii at
Manoa, Honolulu, USA; South Florida Water Manage-
Angelo, A.C. (1997) A galha dos botões do araçazeiro –
Psidium cattleianum Sabine (Myrtaceae), e insetos associados.Curso de Pós-Graduação em Ciências Biológicas.
M.Sc. Dissertation, Universidade Federal do Paraná.
Bennet, F.D. & Habeck, D.H. (1991) Brazilian Peppertree –
Prospects for Biological Control in Florida. Proceedings of
the Symposium on Exotic Pest Plants.
Butignol, C.A. & Pedrosa-Macedo, J.H. (2001) Ocorrência de
parasitóides em Neotrioza tavaresi Crawford (Hemiptera:
Psyllidae), galhador da folha do araçazeiro, Psidium
cattleianum (Myrtaceae). Abstract. Siconbiol VII – Simpósio
de Controle Biológico. p. 321, Poços de Caldas, MG.
Carpanezzi, A.A., Ferreira, C.A., Rotta, E., Namikawa, I.S.,
Sturion, J.A., Pereira, J.C.D., Montagner, L.H., Rauen, M.J.,
Carvalho, P.E.R., Silveira, R.A. & Alves, S.T. (1986)
Zoneamento ecológico para plantios florestais no Estado do
Parana, EMBRAPA-CNPF, Documentos 17.
Diong, C.H. (1982) Population biology and management of the
feral pig (Sus scrofa) in Kipahulu Valley, Maui. Ph. D.
dissertation. University of Hawaii, Honolulu.
Karam, F.S.C., Méndez, M.C., Jarenkow, J.A. & Riet-Correa, F.
(2002) Fenologia de quatro espécies tóxicas de Senecio
(Asteraceae) na região Sul do Rio Grande do Sul. Pesquisa
Veterinária Brasileira 22, 33–39.
Kranz, W.M. & Passini, T. (1997) Amarelinho – Biologia e
Controle. Informe da Sano & Almeida (1998) Pesquisa.
IAPAR, Londrina. XVII (121).
Lee, R.A., Cure, J.R. & Pérez, M.M. (2000) Biodiversity and
relative abundance of insects and spider mites in tree used as
live fences in the Bogotá plateau. Revista Colombiana de
Entomologia 26, 1–15.
Maack, R. (1968) Geografia Física do Estado do Paraná.
CODEPAR, Curitiba, Brazil.
Olckers, T. (1999) Biological control of Solanum mauritianum
Scopoli (Solanaceae) in South Africa: a review of candidate
agents, progress and future prospects. African Entomology
Memoir 1, 65–73.
Pedrosa-Macedo, J.H., Wikler, C., Vitorino, M.D. & Smith,
C.W. (2000) Current researches of Brazilian weeds in
313
Proceedings of the XI International Symposium on Biological Control of Weeds
Paraná state – Biological control of weeds program,
Curitiba, Paraná, Brazil. Proceedings of the X International
Symposium on Biological Control of Weeds (ed N.R.
Spencer), pp.639–643. United States Department of Agriculture, Agricultural Research Services, Sidney, MT and
Montana State University, Bozeman, MT.
Pedrosa-Macedo, J.H. (2001) Convênios e Projetos sobre
Controle Biológico de Plantas – Relatório Anual. FUPEF –
Fundação de Pesquisas Florestais do Paraná. Curitiba, PR,
Brasil.
Roe, K.E. (1972) A revision of Solanum sect. Brevantherum
(Solanaceae). Brittonia 24, 239–278.
Mendonça, R.C., Felfili, J.M., Walter, B.M.T., Silva-Jr., M.C.,
Rezende, A.V., Filgueiras, T. S. & Nogueira, P.E. (1998)
Flora Vascular do Cerrado. Cerrado: Ambiente e Flora. (eds
S.M. Sano& S.P. Almeida), pp. 289–556, Planaltina,
EMBRAPA–CPAC.
Vitorino, M.D. (2001) Aspectos biológicos e testes de especificidade e de reprodução com Heteroperreyia hubrichi
Malaise, 1955 (Hymenoptera: Pergidae) para o controle
biológico da aroeira, Schinus terebinthifolius Raddi (Anacardiaceae) no Estado da Flórida-EUA. Curso de Pós-Graduação em Ciências Biológicas. Doctoral thesis.
Universidedade Federal do Paraná.
Wikler, C. (1999) Distribuição geográfica mundial de Psidium
cattleianum Sabine (Myrtaceae) e um cecidógeno com
possibilidades de utilização em controle biológico. Doctoral
thesis. Universidedade Federal do Paraná.
Workman, R.W. (1978) History of Schinus in Florida. Schinus:
Technical Proceedings of Techniques for Control of Schinus
in South-Florida: A Workshop for Natural Area Managers.
Sanibel-FL, The Sanibel-Captiva Conservation Foundation.
314